Among practices already using or piloting it, daily use is common, but deep integration remains rare. While 60% use AI daily and 95% have it live in at least one workflow, only 7% have reached advanced or fully integrated adoption.

To benchmark where private practices actually stand, Tebra surveyed 292 providers, practice owners, and practice leaders who use or pilot AI, along with 401 patients across the United States. The results mapped what practices run AI for, what they spend, where it pays off, and how comfortable patients are with it in the exam room.

Most adopters had moved past experimentation. Three-fifths (60%) used AI daily, and 68% had adopted it within the past year, while just 7% had reached advanced or fully integrated use.

Depth didn't track neatly with practice size. Here is how it broke down:

Oversight was the weak spot. Fewer than one-tenth of practices (9%) ran AI with no accuracy review, no patient disclosure, and no compliance policy, and 27% had it live with neither a formal compliance policy nor staff training. The gap was widest at the top, where 52% of practice owners lacked a formal compliance policy, compared with 35% of employed providers and 29% of managers.

Documentation is where AI tends to land first, and monthly spending stays modest for most.

Clinical documentation was the leading use, with 74% relying on ambient documentation tools. Other common uses included patient messaging (65%), scheduling (63%), and phone and patient intake (54%).

Most spending stayed restrained. The median AI-using practice spent $100 to $299 per provider each month, though 1 in 5 spent $1,000 or more, and larger groups spent more heavily. Just over one-third of practices (34%) with 11-plus providers topped $1,000 per provider, vs. 10% of 2-to-5-provider practices.

Getting started was fairly quick for many. Nearly one-third (31%) had their main AI tool running in a month or less, and almost half (49%) within 3 months, while 18% were still implementing.

Accuracy was the top barrier to doing more, named by 32% of practices, ahead of compliance or legal concerns (20%) and patient concerns (13%). On the compliance front, 27% had updated patient consent for state AI disclosure laws, and 12% did not know those laws existed.

For practices already using AI, the reported benefits extend beyond documentation, but improvements in day-to-day operations do not always translate into reduced burnout.

Time savings led the benefits. Nearly half of practices (49%) reported less documentation time, cutting it by a third (from 15 hours a week to 10), 38% saw more patient capacity, and 36% reported less after-hours work. Burnout, however, is still prevalent among practices, with only 48% reporting less of it.

Gains in practice revenue climbed with maturity. One-fifth of early adopters (20%) reported more revenue, rising to 32% at routine use and 45% among advanced or fully integrated practices, 2.3 times the early-adopter rate.

The overall verdict was positive, with 73% of practices saying they would recommend adopting AI to a peer.

Patients are open to AI in the background, but they want to be asked when it touches their care.

Consent was close to a universal expectation. Nearly all patients (91%) wanted their permission required every time AI was used for diagnosis or treatment.

Comfort depended on the task. Almost half of patients (49%) were uncomfortable with AI in diagnosis and the same share (49%) in treatment, but only 13% objected to it in scheduling. Discomfort with AI diagnosis was highest among the youngest patients, with 67% of Gen Z uncomfortable, compared with 36% of Baby Boomers.

For your practice, that gap points to an important step. Build clear disclosure and consent into any patient-facing AI, especially anything that touches diagnosis or treatment, and keep the lighter-touch uses like scheduling running in the background.

AI adoption is no longer the defining challenge for practices already using the technology. The next test is whether they can integrate it into everyday operations, demonstrate meaningful returns and establish safeguards that earn patient trust. Practices with more mature AI use report stronger revenue outcomes, but the findings also reveal gaps in oversight and patient expectations that technology alone cannot resolve. 

Tebra commissioned a 2-part survey to benchmark how private and small medical practices adopt AI and how patients feel about AI in their care. Fieldwork ran across 2 instruments in 2026: a provider survey of 292 private providers, practice owners, and practice leaders, and a patient survey of 401 US adults. Provider respondents were screened to those currently using or piloting AI in clinical or administrative work. Both samples were fielded through CloudResearch Connect and Prolific.

These are nonprobability opt-in samples, so should be interpreted  as directional and are not statistically projectable to all private practices or U.S. patients. Conventional probability-sampling margins of error are not applicable.  rather than projectable to all private practices or all patients. The margin of error is approximately plus or minus 5.7 percentage points for the provider sample and plus or minus 4.9 points for the patient sample at a 95% confidence level. 

Percentages are rounded to whole numbers and may not total 100. Select-all-that-apply items can exceed 100%. Numeric responses report an average with outliers removed using the interquartile range method. Age was collected as birth year and grouped into generations. Demographic subgroups representing under 5% of a sample are treated as directional only. Combined figures are calculated from unrounded data, so they may differ by a point from the sum of individually rounded figures shown in charts.

The Practice AI Maturity Model classifies each practice on observable behaviors rather than self-assessment, combining the number of workflows where AI is live with 4 governance signals: consistent accuracy review, consistent patient disclosure, a formal AI and HIPAA compliance policy, and a formal staff training program.

Provider sample (n = 292)

Patient sample (n = 401)

Tebra, headquartered in Southern California, empowers private healthcare practices with AI and automation to drive growth, streamline care, and boost efficiency. Our all-in-one EHR and billing platform delivers everything you need to attract and engage your patients, including online scheduling, reputation management, and digital communications.

Inspired by "vertebrae," our name embodies our mission to be the backbone of healthcare success. With over 165,000 providers and 190 million patient records, Tebra is redefining healthcare through innovation and a commitment to customer success. We're not just optimizing operations; we're ensuring private practices thrive.

You're welcome to share these findings for noncommercial purposes. When you do, please link back to this page so readers can see the full study and methodology, and credit Tebra as the source.